Make all figures in this notebook to keep the results reproducible and the styling consistent.
Created: 4/11/2021
Author: Andrew Weng
import os, sys
if os.path.basename(os.getcwd()) == 'code-base':
os.chdir('../')
sys.path.append('code-base')
import numpy as np
import pandas as pd
import glob, re, pprint, random
from datetime import datetime
import pprint
from scipy.stats import pearsonr, ttest_ind
from scipy.signal import savgol_filter
from scipy import interpolate
from sklearn.linear_model import LinearRegression
import seaborn as sns
from matplotlib import pyplot as plt
from src.formation import FormationCell
from src.utils import get_label_registry
import warnings
warnings.filterwarnings('ignore')
assert os.path.basename(os.getcwd()) == 'project-formation'
pd.set_option('display.max_columns', 500)
pd.set_option('display.max_rows', 50)
# Configure plot settings
plt.rc('font', **{'family' : 'sans-serif',
'sans-serif' : ['Arial'],
'size': 18
})
plt.rc('figure', **{'autolayout' : True,
'figsize' : (7, 5),
'facecolor' : 'white',
'dpi' : 150
})
plt.rc('xtick', labelsize='medium')
plt.rc('ytick', labelsize='medium')
plt.rc('axes', labelsize='medium', grid=False)
plt.rc('axes', titlesize='medium')
plt.rc('legend', fontsize='medium')
# Formation types
CELL_BASE = 7
CELL_FAST = 33
COLOR_BASE = np.array([0, 0, 0])
COLOR_FAST = np.array([44, 121, 245])/255
COLOR_FAST_RT = np.array([0, 0, 1])
COLOR_FAST_HT = np.array([1, 0, 0])
LINESTYLE_BASE = ':'
LINESTYLE_FAST = '-'
# Pos/neg electrode colors
COLOR_POS = np.array([0, 0, 1])
COLOR_NEG = np.array([1, 0, 0])
COLOR_FULL_CELL = np.array([0, 0, 0])
COLOR_REF = (0.7, 0.7, 0.7)
NOMINAL_CAPACITY_AH = 2.36
CAPACITY_LIMITS_AH = (0.8, 2.4)
TARGET_RETENTION = 0.7
cell_dict = dict()
cellid_list = np.arange(1, 41)
for cellid in cellid_list:
cell_dict[cellid] = FormationCell(cellid)
df_base_form = cell_dict[CELL_BASE].get_formation_data()
df_fast_form = cell_dict[CELL_FAST].get_formation_data()
df_base_form = df_base_form[df_base_form['Test Time (s)'] > 86400]
df_fast_form = df_fast_form[df_fast_form['Test Time (s)'] > 86400]
plt.figure()
plt.plot((df_base_form['Test Time (s)'] - 86400)/3600, df_base_form['Potential (V)'],
color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot((df_fast_form['Test Time (s)'] - 86400)/3600, df_fast_form['Potential (V)'],
color=COLOR_FAST, linestyle=LINESTYLE_FAST)
plt.xlabel('Time (hrs)')
plt.ylabel('Voltage (V)')
lgd = plt.legend(['Baseline Formation', 'Fast Formation'], frameon=True)
plt.ylim((2.9, 4.3))
plt.xlim((-1, 65))
(-1.0, 65.0)
df_base_rpt = cell_dict[CELL_BASE].get_aging_data_timeseries()
df_base_rpt = df_base_rpt[df_base_rpt['Cycle Number'] < 5]
df_base_rpt = df_base_rpt[df_base_rpt['Cycle Number'] > 1]
df_fast_rpt = cell_dict[CELL_FAST].get_aging_data_timeseries()
df_fast_rpt = df_fast_rpt[df_fast_rpt['Cycle Number'] < 5]
df_fast_rpt = df_fast_rpt[df_fast_rpt['Cycle Number'] > 1]
plt.subplot()
plt.plot(df_base_rpt['Test Time (s)'] / 3600, df_base_rpt['Potential (V)'],
color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot(df_fast_rpt['Test Time (s)'] / 3600, df_fast_rpt['Potential (V)'],
color=COLOR_FAST, linestyle=LINESTYLE_FAST)
plt.xlabel('Time (hrs)')
plt.ylabel('Voltage (V)')
plt.ylim((2.9, 4.3))
plt.legend(['Baseline', 'Fast'], frameon=False)
<matplotlib.legend.Legend at 0x135fcf640>
df_base_cyc = cell_dict[CELL_BASE].get_aging_data_cycles()
idx = np.where((df_base_cyc['Total Charge Time (s)'] > 8500) |
(df_base_cyc['Total Charge Time (s)'] < 100))[0]
df_base_cyc['Discharge Capacity (Ah)'][idx] = np.nan
df_fast_cyc = cell_dict[CELL_FAST].get_aging_data_cycles()
idx = np.where((df_fast_cyc['Total Charge Time (s)'] > 8500) |
(df_fast_cyc['Total Charge Time (s)'] < 100))[0]
df_fast_cyc['Discharge Capacity (Ah)'][idx] = np.nan
plt.figure()
plt.plot(df_base_cyc['Cycle Number'], df_base_cyc['Discharge Capacity (Ah)'],
color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot(df_fast_cyc['Cycle Number'], df_fast_cyc['Discharge Capacity (Ah)'],
color=COLOR_FAST, linestyle=LINESTYLE_FAST)
plt.axhline(y=TARGET_RETENTION*NOMINAL_CAPACITY_AH , color='r', linestyle='--')
plt.xlabel('Cycle Number')
plt.ylabel('Discharge Capacity (Ah)')
plt.legend(['Baseline', 'Fast'], frameon=False)
plt.ylim(CAPACITY_LIMITS_AH)
plt.xlim((0, 500))
(0.0, 500.0)
# Process voltage decay signal during 12-hour rest step
STEP_INDEX_6HR_REST = 12
CYCLE_INDEX_6HR_REST = 3
VOLTAGE_MAXIMUM = 4.2
df_base_deltav = df_base_form[(df_base_form['Cycle Number'] == 3) &
(
(df_base_form['Step Index'] == 11) |
(df_base_form['Step Index'] == 12)
)]
df_base_deltav['Test Time (s)'] -= np.min(df_base_deltav['Test Time (s)'])
df_fast_deltav = df_fast_form[(df_fast_form['Cycle Number'] == 7) &
(
(df_fast_form['Step Index'] == 13) |
(df_fast_form['Step Index'] == 12)
)]
df_fast_deltav['Test Time (s)'] -= np.min(df_fast_deltav['Test Time (s)'])
plt.figure(figsize=(5, 4))
plt.plot((df_base_deltav['Test Time (s)'])/3600, df_base_deltav['Potential (V)'],
color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot((df_fast_deltav['Test Time (s)'])/3600, df_fast_deltav['Potential (V)'],
color=COLOR_FAST, linestyle=LINESTYLE_FAST)
plt.xlabel('Time (hrs)')
plt.ylabel('Voltage (V)')
lgd = plt.legend(['Baseline', 'Fast'], frameon=False)
plt.ylim((4.175, 4.205))
plt.xlim((0, 6.7))
(0.0, 6.7)
plt.figure(figsize=(6, 5))
plt.plot(np.nan, np.nan, label='Fast', color=COLOR_FAST)
plt.plot(np.nan, np.nan, label='Baseline', color=COLOR_BASE)
for cell in cell_dict.values():
if cell.cellid == 9:
continue
if random.random() > 0.2:
continue
df_form = cell.get_formation_data()
if cell.is_baseline_formation():
color = COLOR_BASE
linestyle = LINESTYLE_FAST
df_deltav = df_form[(df_form['Cycle Number'] == 3) &
(
(df_form['Step Index'] == 11) |
(df_form['Step Index'] == 12)
)]
# df_deltav['Potential (V)'][df_deltav['Step Index'] == 12] = \
# savgol_filter(df_deltav['Potential (V)'][df_deltav['Step Index'] == 12], 1, 3)
else:
color = COLOR_FAST
linestyle = LINESTYLE_FAST
df_deltav = df_form[(df_form['Cycle Number'] == 7) &
(
(df_form['Step Index'] == 13) |
(df_form['Step Index'] == 12)
)]
# df_deltav['Potential (V)'][df_deltav['Step Index'] == 13] = \
# savgol_filter(df_deltav['Potential (V)'][df_deltav['Step Index'] == 13], 1, 3)
df_deltav['Test Time (s)'] -= np.min(df_deltav['Test Time (s)'])
plt.plot((df_deltav['Test Time (s)'])/3600, df_deltav['Potential (V)'],
color=color,
linestyle=linestyle,
linewidth=1,
label='',
alpha=1)
plt.xlabel('Time (hrs)')
plt.ylabel('Voltage (V)')
lgd = plt.legend(frameon=False)
hppc_base_list = cell_dict[CELL_BASE].process_diagnostic_hppc_data()
# Unpack data from first available cycle
df_hppc_base = hppc_base_list[0]['data']
hppc_fast_list = cell_dict[CELL_FAST].process_diagnostic_hppc_data()
# Unpack data from first available cycle
df_hppc_fast = hppc_fast_list[0]['data']
df_hppc_fast_aged = hppc_fast_list[2]['data']
plt.figure(figsize=(9,4.5))
# plt.plot(df_hppc_base['capacity'],
# df_hppc_base['resistance_10s_ohm']*1000,
# color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot(df_hppc_fast['capacity'],
df_hppc_fast['resistance_10s_ohm']*1000,
color='k', linestyle=':')
plt.plot(df_hppc_fast_aged['capacity'],
df_hppc_fast_aged['resistance_10s_ohm']*1000,
color='k', linestyle=LINESTYLE_FAST)
plt.xlabel('Capacity (Ah)')
plt.ylabel('$R_{10s}$ ($m\Omega$)')
plt.xlim((-0.3, 3))
# lgd = plt.legend(['Baseline', 'Fast'], frameon=False)
(-0.3, 3.0)
plt.figure()
plt.plot(df_hppc_base['capacity']/np.max(df_hppc_base['capacity'])*100,
df_hppc_base['resistance_10s_ohm'],
color=COLOR_BASE, linestyle=LINESTYLE_BASE, marker='o')
plt.plot(df_hppc_fast['capacity']/np.max(df_hppc_fast['capacity'])*100,
df_hppc_fast['resistance_10s_ohm'],
color=COLOR_FAST, linestyle=LINESTYLE_FAST, marker='o')
plt.xlabel('SOC (%)')
plt.ylabel('$R_{10s}$ ($\Omega$)')
plt.xlim((0, 20))
lgd = plt.legend(['Baseline', 'Fast'], frameon=False)
c20_base_list = cell_dict[CELL_BASE].process_diagnostic_c20_data()
c20_fast_list = cell_dict[CELL_FAST].process_diagnostic_c20_data()
df_c20_base = c20_base_list[0]
df_c20_fast = c20_fast_list[0]
plt.figure()
plt.plot(df_c20_base['chg_capacity']/np.max(df_c20_fast['chg_capacity'])*100,
df_c20_base['chg_dvdq'], linestyle=LINESTYLE_BASE, color=COLOR_BASE)
plt.plot(df_c20_fast['chg_capacity']/np.max(df_c20_fast['chg_capacity'])*100,
df_c20_fast['chg_dvdq'], linestyle=LINESTYLE_FAST, color=COLOR_FAST)
plt.ylim((0, 0.6))
plt.xlabel('SOC (%)')
plt.ylabel('dV/dQ (V/Ah)')
plt.legend(['Baseline', 'Fast'], frameon=False)
<matplotlib.legend.Legend at 0x138f318e0>
cellid_list = np.arange(1, 41)
num_base_cells = 20
num_fast_cells = 20
cyc_number_interp = np.arange(1, 501, 1)
cap_mat_fast = np.empty((len(cyc_number_interp), num_fast_cells))
cap_mat_fast[:, :] = np.NaN
idx_fast = 0
cap_mat_base = np.empty((len(cyc_number_interp), num_base_cells))
cap_mat_base[:, :] = np.NaN
idx_base = 0
for idx_cell, cellid in enumerate(cellid_list):
cell = FormationCell(cellid)
if cell.is_room_temp() or cell.cellid == 9:
continue
df = cell.get_aging_data_cycles()
idx = np.where((df['Total Charge Time (s)'] > 8500) |
(df['Total Charge Time (s)'] < 100))[0]
df = df.copy()
df['Discharge Capacity (Ah)'][idx] = np.nan
# Remove some things from the back
cyc = df['Cycle Number']
cap = df['Discharge Capacity (Ah)']
cyc = cyc[~np.isnan(cap)]
cap = cap[~np.isnan(cap)]
cap_interp = np.interp(cyc_number_interp, cyc, cap, left=np.NaN, right=np.NaN)
if cell.is_baseline_formation():
color = COLOR_BASE
linestyle = ':'
cap_mat_base[:, idx_base] = cap_interp
idx_base += 1
else:
color = COLOR_FAST_HT
linestyle = ':'
cap_mat_fast[:, idx_fast] = cap_interp
idx_fast += 1
plt.plot(cyc_number_interp, cap_interp, color=color, linestyle=linestyle)
cap_mean_base = np.nanmean(cap_mat_base, axis=1)
cap_mean_fast = np.nanmean(cap_mat_fast, axis=1)
cap_std_base = np.nanstd(cap_mat_base, axis=1)
cap_std_fast = np.nanstd(cap_mat_fast, axis=1)
plt.plot(cyc_number_interp, cap_mean_base, linewidth=3, color=COLOR_BASE, label='Baseline')
plt.plot(cyc_number_interp, cap_mean_fast, linewidth=3, color=COLOR_FAST_HT, label='Fast')
plt.xlabel('Cycle Number')
plt.ylabel('Discharge Capacity (Ah)')
plt.legend()
plt.figure()
plt.plot(cyc_number_interp, cap_std_base/cap_mean_base, linewidth=3, color=COLOR_BASE, label='Baseline')
plt.plot(cyc_number_interp, cap_std_fast/cap_mean_fast, linewidth=3, color=COLOR_FAST_HT, label='Fast')
plt.xlabel('Cycle Number')
plt.ylabel('Coefficient of Variation, $\sigma / \mu$')
plt.legend()
<matplotlib.legend.Legend at 0x138710fa0>
import scipy
print(scipy.version.version)
1.7.0
cellid_list = np.arange(1, 41)
plt.figure(figsize=(10, 4.5))
for cellid in cellid_list:
cell = FormationCell(cellid)
if cell.is_room_temp() or cell.cellid == 9:
continue
if cell.is_baseline_formation():
color = COLOR_BASE
linestyle = LINESTYLE_BASE
else:
color = COLOR_FAST_HT
linestyle = LINESTYLE_FAST
df = cell.get_aging_data_cycles()
idx = np.where((df['Total Charge Time (s)'] > 8500) |
(df['Total Charge Time (s)'] < 100))[0]
df = df.copy()
df['Discharge Capacity (Ah)'][idx] = np.nan
cyc = df['Cycle Number']
cap = df['Discharge Capacity (Ah)']
plt.plot(cyc, cap,
color=color, linestyle=linestyle)
plt.plot(np.nan, np.nan, label='Baseline ',
color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot(np.nan, np.nan, label='Fast ',
color=COLOR_FAST_HT, linestyle=LINESTYLE_FAST)
plt.axhline(y=TARGET_RETENTION*NOMINAL_CAPACITY_AH,
color=COLOR_REF, linestyle='--')
plt.legend(loc=3, frameon=False)
plt.xlabel('Cycle Number')
plt.ylabel('Discharge Capacity (Ah)')
plt.ylim(CAPACITY_LIMITS_AH)
(0.8, 2.4)
plt.figure(figsize=(10, 4.5))
for cellid in cellid_list:
cell = FormationCell(cellid)
if not cell.is_room_temp() or cell.cellid == 9:
continue
if cell.is_baseline_formation():
color = COLOR_BASE
linestyle = LINESTYLE_BASE
else:
color = COLOR_FAST_RT
linestyle = LINESTYLE_FAST
df = cell.get_aging_data_cycles()
idx = np.where((df['Total Charge Time (s)'] > 8500) |
(df['Total Charge Time (s)'] < 100))[0]
df['Discharge Capacity (Ah)'][idx] = np.nan
cyc = df['Cycle Number']
cap = df['Discharge Capacity (Ah)']
plt.plot(cyc, cap,
color=color, linestyle=linestyle)
plt.plot(np.nan, np.nan, label='Baseline Formation',
color=COLOR_BASE, linestyle=LINESTYLE_BASE)
plt.plot(np.nan, np.nan, label='Fast Formation',
color=COLOR_FAST_RT, linestyle=LINESTYLE_FAST)
plt.legend(loc=3, frameon=False)
plt.axhline(y=TARGET_RETENTION*NOMINAL_CAPACITY_AH,
color=COLOR_REF, linestyle='--')
plt.xlabel('Cycle Number')
plt.ylabel('Discharge Capacity (Ah)')
plt.ylim(CAPACITY_LIMITS_AH)
(0.8, 2.4)
variable_name = 'cycles_to_70_pct'
df = pd.read_csv('output/correlation_data.csv')
group1 = df[(df['is_baseline_formation'] == False) & (df['is_room_temp_aging'] == True)][variable_name]
group2 = df[(df['is_baseline_formation'] == True) & (df['is_room_temp_aging'] == True)][variable_name]
# group1 = df[(df['is_baseline_formation'] == False)][variable_name]
# group2 = df[(df['is_baseline_formation'] == True)][variable_name]
all_data = [group1, group2]
labels = ['Fast', 'Baseline']
fig, (ax1) = plt.subplots(nrows=1, ncols=1, figsize=(5, 4))
# rectangular box plot
bplot = ax1.boxplot(all_data,
vert=True, # vertical box alignment
patch_artist=True, # fill with color
labels=labels,
showfliers=False,
widths=0.6) # will be used to label x-ticks
x = np.random.normal(1, 0.06, size=len(all_data[0]))
plt.plot(x, all_data[0], marker='o',
markersize=4, linestyle='', color=COLOR_FAST_RT, alpha=0.7)
x = np.random.normal(2, 0.08, size=len(all_data[1]))
plt.plot(x, all_data[1], marker='o',
markersize=4, linestyle='', color=COLOR_BASE, alpha=0.7)
# fill with colors
colors = [COLOR_FAST_RT, COLOR_BASE]
for patch, color in zip(bplot['boxes'], colors):
patch.set_facecolor(color)
patch.set_alpha(0.2)
# adding horizontal grid lines
for ax in [ax1]:
ax.yaxis.grid(False)
ax.set_ylabel('Cycles to 70%')
ax.set_ylim((300, 600))
plt.show()
fig, axs = plt.subplots(2, 1, figsize=(5, 10))
for cellid in cellid_list:
cell = FormationCell(cellid)
if cell.cellid == 9:
continue
# if cellid != CELL_BASE and cellid != CELL_FAST and cellid != 26 and cellid != 15:
# continue
print(f"Working on cell {cellid}...")
ax = axs[0] if cell.is_room_temp() else axs[1]
hppc_dict = cell.process_diagnostic_hppc_data()
cap = hppc_dict[0]['data']['capacity']
dcr = hppc_dict[0]['data']['resistance_10s_ohm']*1000
soc = cap/np.max(cap)*100
color = (0, 0, 1) if cell.is_room_temp() else (1, 0, 0)
color = (0, 0, 0) if cell.is_baseline_formation() else color
ax.plot(soc, dcr, color=color, marker='o', markersize=3, alpha=0.70)
# ax.set_xticks([1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 50, 100])
# ax.set_xlim((3, 10))
[ax.set(ylabel='$R_{10s}$ (m$\Omega$)') for ax in axs]
[ax.set(xlabel='SOC (%)') for ax in axs]
[ax.grid(False) for ax in axs]
plt.show()
Working on cell 1... Working on cell 2... Working on cell 3... Working on cell 4... Working on cell 5... Working on cell 6... Working on cell 7... Working on cell 8... Working on cell 10... Working on cell 11... Working on cell 12... Working on cell 13... Working on cell 14... Working on cell 15... Working on cell 16... Working on cell 17... Working on cell 18... Working on cell 19... Working on cell 20... Working on cell 21... Working on cell 22... Working on cell 23... Working on cell 24... Working on cell 25... Working on cell 26... Working on cell 27... Working on cell 28... Working on cell 29... Working on cell 30... Working on cell 31... Working on cell 32... Working on cell 33... Working on cell 34... Working on cell 35... Working on cell 36... Working on cell 37... Working on cell 38... Working on cell 39... Working on cell 40...
Plot the evolution of x_0, x_100, y_0, y_100 over time
plt.figure(figsize=(6, 12))
ax1 = plt.subplot(511)
ax2 = plt.subplot(512)
ax3 = plt.subplot(513)
ax4 = plt.subplot(514)
ax5 = plt.subplot(515)
for cellid in cellid_list:
cell = cell_dict[cellid]
if cell.is_room_temp() or cell.cellid == 9:
continue
if cell.is_baseline_formation():
color = COLOR_BASE
else:
color = COLOR_FAST
esoh_data_list = cell_dict[cellid].get_esoh_fitting_data()
cyc_number_list = [x['cycle_index'] for x in esoh_data_list]
x_0_list = [x['x0'] for x in esoh_data_list]
y_0_list = [x['y0'] for x in esoh_data_list]
x_100_list = [x['x100'] for x in esoh_data_list]
y_100_list = [x['y100'] for x in esoh_data_list]
rmse_list = [x['RMSE_mV'] for x in esoh_data_list]
ax1.plot(cyc_number_list, x_0_list, color=color)
ax1.set_ylabel('$x_0$')
ax1.set_ylim((-0.01, 0.02))
ax2.plot(cyc_number_list, y_0_list, color=color)
ax2.set_ylabel('$y_0$')
ax2.set_ylim((0.85, 0.98))
ax3.plot(cyc_number_list, x_100_list, color=color)
ax3.set_ylabel('$x_{100}$')
ax4.plot(cyc_number_list, y_100_list, color=color)
ax4.set_ylabel('$y_{100}$')
ax5.plot(cyc_number_list, rmse_list, color=color)
ax5.set_ylabel('$RMSE_{mV}$')
ax5.set_xlabel('Cycle Number')
ax1.plot(np.nan, np.nan, color=COLOR_BASE, label='Base')
ax1.plot(np.nan, np.nan, color=COLOR_FAST, label='Fast')
ax1.legend()
<matplotlib.legend.Legend at 0x13ae96e50>
Prepare the two datasets for comparison
Compare the same cell, two aging conditions
esoh_data_1 = cell_dict[CELL_FAST].get_esoh_fitting_data()[0]
esoh_data_2 = cell_dict[CELL_FAST].get_esoh_fitting_data()[3]
print(f"Cycle {esoh_data_1['cycle_index']} to {esoh_data_2['cycle_index']}")
print(f"y0: {esoh_data_1['y0']:.3f} --> {esoh_data_2['y0']:.3f}")
print(f"x0: {esoh_data_1['x0']:.3f} --> {esoh_data_2['x0']:.3f}")
print(f"y100: {esoh_data_1['y100']:.3f} --> {esoh_data_2['y100']:.3f}")
print(f"x100: {esoh_data_1['x100']:.3f} --> {esoh_data_2['x100']:.3f}")
print(f"---")
print(f"LAM_PE = {100 - 100 * esoh_data_2['Cp'] / esoh_data_1['Cp']:.1f}%")
print(f"LAM_NE = {100 - 100 * esoh_data_2['Cn'] / esoh_data_1['Cn']:.1f}%")
print(f"LLI = {100 * (esoh_data_1['n_li'] - esoh_data_2['n_li']) / esoh_data_1['n_li'] :.1f}%")
Cycle 3.0 to 262.0 y0: 0.950 --> 0.902 x0: 0.007 --> -0.001 y100: 0.038 --> 0.031 x100: 0.922 --> 0.829 --- LAM_PE = 5.8% LAM_NE = 0.9% LLI = 11.3%
Here is another dataset for comparison
Compare a baseline vs fast formation cell, both fresh
esoh_data_1 = cell_dict[CELL_BASE].get_esoh_fitting_data()[0]
esoh_data_2 = cell_dict[CELL_FAST].get_esoh_fitting_data()[0]
print(f" Base --> Fast")
print(f"y0: {esoh_data_1['y0']:.3f} --> {esoh_data_2['y0']:.3f}")
print(f"x0: {esoh_data_1['x0']:.3f} --> {esoh_data_2['x0']:.3f}")
print(f"y100: {esoh_data_1['y100']:.3f} --> {esoh_data_2['y100']:.3f}")
print(f"x100: {esoh_data_1['x100']:.3f} --> {esoh_data_2['x100']:.3f}")
print(f"---")
print(f"LAM_PE = {100 - 100 * esoh_data_2['Cp'] / esoh_data_1['Cp']:.1f}%")
print(f"LAM_NE = {100 - 100 * esoh_data_2['Cn'] / esoh_data_1['Cn']:.1f}%")
print(f"LLI = {100 * (esoh_data_1['n_li'] - esoh_data_2['n_li']) / esoh_data_1['n_li'] :.1f}%")
Base --> Fast y0: 0.954 --> 0.950 x0: 0.009 --> 0.007 y100: 0.037 --> 0.038 x100: 0.905 --> 0.922 --- LAM_PE = 0.2% LAM_NE = 2.8% LLI = 0.8%
xlim_array = [(-0.3, 3),
(-0.3, 0.3),
(2.0, 2.6)]
neg_ylim_array = [(-0.05, 1.05),
(-0.05, 1.05),
(0.08, 0.095)]
pos_ylim_array = [(2.4, 4.55),
(2.95, 3.75),
(4.10, 4.39)]
# dV/dQ
fig, ax = plt.subplots(figsize=(9,6))
plt.plot(esoh_data_1['pos']['Q'], esoh_data_1['pos']['dVdQ'], color=COLOR_POS, linestyle=':')
plt.plot(esoh_data_1['orig']['Q'], esoh_data_1['orig']['dVdQ'], color=COLOR_FULL_CELL, linestyle=':')
plt.plot(esoh_data_2['pos']['Q'], esoh_data_2['pos']['dVdQ'], color=COLOR_POS)
plt.plot(esoh_data_2['orig']['Q'], esoh_data_2['orig']['dVdQ'], color=COLOR_FULL_CELL)
plt.plot(esoh_data_1['neg']['Q'], esoh_data_1['neg']['dVdQ'], color=COLOR_NEG, linestyle=':')
plt.plot(esoh_data_2['neg']['Q'], esoh_data_2['neg']['dVdQ'], color=COLOR_NEG)
plt.ylim((0, 0.75))
plt.xlabel('Capacity (Ah)')
plt.ylabel('dV/dQ (V/Ah)')
Text(0, 0.5, 'dV/dQ (V/Ah)')
esoh_data_1 = cell_dict[CELL_FAST].get_esoh_fitting_data()[0]
esoh_data_2 = cell_dict[CELL_FAST].get_esoh_fitting_data()[2]
esoh_data_3 = cell_dict[CELL_BASE].get_esoh_fitting_data()[0]
esoh_data_4 = cell_dict[CELL_BASE].get_esoh_fitting_data()[3]
print(f" Base --> Fast --> Fast (aged)")
print(f"y0: {esoh_data_3['y0']:.3f} --> {esoh_data_1['y0']:.3f} --> {esoh_data_2['y0']:.3f}")
print(f"x0: {esoh_data_3['x0']:.3f} --> {esoh_data_1['x0']:.3f} --> {esoh_data_2['x0']:.3f}")
print(f"y100: {esoh_data_3['y100']:.3f} --> {esoh_data_1['y100']:.3f} --> {esoh_data_2['y100']:.3f}")
print(f"x100: {esoh_data_3['x100']:.3f} --> {esoh_data_1['x100']:.3f} --> {esoh_data_2['x100']:.3f}")
print(f"---")
print(f"Fast --> Fast (aged)")
print(f"LAM_PE = {100 - 100 * esoh_data_2['Cp'] / esoh_data_1['Cp']:.1f}%")
print(f"LAM_NE = {100 - 100 * esoh_data_2['Cn'] / esoh_data_1['Cn']:.1f}%")
print(f"LLI = {100 * (esoh_data_1['n_li'] - esoh_data_2['n_li']) / esoh_data_1['n_li'] :.1f}%")
print(f"---")
print(f"Base --> Fast")
print(f"LAM_PE = {100 - 100 * esoh_data_1['Cp'] / esoh_data_3['Cp']:.1f}%")
print(f"LAM_NE = {100 - 100 * esoh_data_1['Cn'] / esoh_data_3['Cn']:.1f}%")
print(f"LLI = {100 * (esoh_data_3['n_li'] - esoh_data_1['n_li']) / esoh_data_3['n_li'] :.1f}%")
print(f"Base --> Base (aged)")
print(f"LAM_PE = {100 - 100 * esoh_data_4['Cp'] / esoh_data_3['Cp']:.1f}%")
print(f"LAM_NE = {100 - 100 * esoh_data_4['Cn'] / esoh_data_3['Cn']:.1f}%")
print(f"LLI = {100 * (esoh_data_3['n_li'] - esoh_data_4['n_li']) / esoh_data_3['n_li'] :.1f}%")
Base --> Fast --> Fast (aged) y0: 0.954 --> 0.950 --> 0.918 x0: 0.009 --> 0.007 --> -0.000 y100: 0.037 --> 0.038 --> 0.032 x100: 0.905 --> 0.922 --> 0.847 --- Fast --> Fast (aged) LAM_PE = 4.4% LAM_NE = -0.2% LLI = 8.4% --- Base --> Fast LAM_PE = 0.2% LAM_NE = 2.8% LLI = 0.8% Base --> Base (aged) LAM_PE = 6.4% LAM_NE = 2.4% LLI = 12.0%
fig_arr, ax_arr, ax2_arr = [], [], []
for (xlim, neg_ylim, pos_ylim) in zip(xlim_array, neg_ylim_array, pos_ylim_array):
fig, ax = plt.subplots(figsize=(9,5.5))
ax2 = ax.twinx()
fig_arr.append(fig)
ax_arr.append(ax)
ax2_arr.append(ax2)
ax.plot(esoh_data_1['pos']['Q'], esoh_data_1['pos']['V'], color=COLOR_POS, linestyle=':')
ax.plot(esoh_data_1['orig']['Q'], esoh_data_1['orig']['V'], color=COLOR_FULL_CELL, linestyle=':')
ax2.plot(esoh_data_1['neg']['Q'], esoh_data_1['neg']['V'], color=COLOR_NEG, linestyle='-')
ax.plot(esoh_data_2['pos']['Q'], esoh_data_2['pos']['V'], color=COLOR_POS)
ax.plot(esoh_data_2['orig']['Q'], esoh_data_2['orig']['V'], color=COLOR_FULL_CELL)
# ax2.plot(esoh_data_2['neg']['Q'], esoh_data_2['neg']['V'], color=COLOR_NEG)
# ax.plot(esoh_data_3['pos']['Q'], esoh_data_3['pos']['V'], color=np.array([3, 211, 252])/255, linestyle=':')
# ax.plot(esoh_data_3['orig']['Q'], esoh_data_3['orig']['V'], color=np.array([0.5, 0.5, 0.5]), linestyle=':')
# ax2.plot(esoh_data_3['neg']['Q'], esoh_data_3['neg']['V'], color=np.array([252, 123, 3])/255, linestyle=':')
# ax.plot(esoh_data_4['pos']['Q'], esoh_data_4['pos']['V'], color=np.array([3, 211, 252])/255, linestyle='-')
# ax.plot(esoh_data_4['orig']['Q'], esoh_data_4['orig']['V'], color=np.array([0.5, 0.5, 0.5]), linestyle='-')
# ax2.plot(esoh_data_4['neg']['Q'], esoh_data_4['neg']['V'], color=np.array([252, 123, 3])/255, linestyle='-')
ax.set_xlabel('Capacity (Ah)')
ax2.yaxis.label.set_color(COLOR_NEG)
ax2.spines["right"].set_edgecolor(COLOR_NEG)
ax2.tick_params(axis='y', colors=COLOR_NEG)
ax.set_ylabel('V / V vs $Li/Li^+$')
ax2.set_ylabel('V vs $Li/Li^+$')
ax.set_xlim(xlim)
ax2.set_ylim(neg_ylim)
ax.set_ylim(pos_ylim)
# ax_arr[1].plot([0, 0], [0, 5], color=COLOR_REF, linestyle='--')
#
# ax_arr[2].plot([esoh_data_1['orig']['Q'][-1], esoh_data_1['orig']['Q'][-1]],
# [0, 5], color=COLOR_REF, linestyle='--')
# ax_arr[2].plot([esoh_data_2['orig']['Q'][-1], esoh_data_2['orig']['Q'][-1]],
# [0, 5], color=COLOR_REF, linestyle='--')
# ax_arr[2].plot([esoh_data_3['orig']['Q'][-1], esoh_data_3['orig']['Q'][-1]],
# [0, 5], color=COLOR_REF, linestyle='--')
# ax_arr[2].plot([esoh_data_4['orig']['Q'][-1], esoh_data_4['orig']['Q'][-1]],
# [0, 5], color=COLOR_REF, linestyle='--')
# dV/dQ
fig, ax = plt.subplots(figsize=(9,6))
for cellid in cellid_list:
cell = cell_dict[cellid]
if cell.is_room_temp() or cell.cellid == 9:
continue
if cell.is_baseline_formation():
linestyle = ':'
linewidth = 0.5
else:
linestyle = '-'
linewidth = 1
esoh_data = cell.get_esoh_fitting_data()[0]
assert esoh_data['cycle_index'] == 3
plt.plot(esoh_data['pos']['Q'], esoh_data['pos']['dVdQ'],
color=COLOR_POS, linestyle=linestyle, linewidth=linewidth)
plt.plot(esoh_data['orig']['Q'], esoh_data['orig']['dVdQ'],
color=COLOR_FULL_CELL, linestyle=linestyle, linewidth=linewidth)
plt.plot(esoh_data['neg']['Q'], esoh_data['neg']['dVdQ'],
color=COLOR_NEG, linestyle=linestyle, linewidth=linewidth)
plt.ylim((0, 0.75))
plt.xlabel('Capacity (Ah)')
plt.ylabel('dV/dQ (V/Ah)')
Text(0, 0.5, 'dV/dQ (V/Ah)')
data = cell.get_esoh_fitting_data()
[x['cycle_index'] for x in data]
[3.0, 56.0, 159.0, 262.0, 365.0, 468.0]
fig_arr, ax_arr, ax2_arr = [], [], []
for (xlim, neg_ylim, pos_ylim) in zip(xlim_array, neg_ylim_array, pos_ylim_array):
fig, ax = plt.subplots(figsize=(9,6))
ax2 = ax.twinx()
fig_arr.append(fig)
ax_arr.append(ax)
ax2_arr.append(ax2)
for cellid in cellid_list:
cell = cell_dict[cellid]
if cell.is_room_temp() or cell.cellid == 9:
continue
if cell.is_baseline_formation():
linestyle = '--'
linewidth = 0.5
else:
linestyle = '-'
linewidth = 1
esoh_data = cell.get_esoh_fitting_data()[0]
assert esoh_data['cycle_index'] == 3
ax.plot(esoh_data['pos']['Q'], esoh_data['pos']['V'],
color=COLOR_POS, linestyle=linestyle, linewidth=linewidth)
ax.plot(esoh_data['orig']['Q'], esoh_data['orig']['V'],
color=COLOR_FULL_CELL, linestyle=linestyle, linewidth=linewidth)
ax2.plot(esoh_data['neg']['Q'], esoh_data['neg']['V'],
color=COLOR_NEG, linestyle=linestyle, linewidth=linewidth)
ax.set_xlabel('Capacity (Ah)')
ax2.yaxis.label.set_color(COLOR_NEG)
ax2.spines["right"].set_edgecolor(COLOR_NEG)
ax2.tick_params(axis='y', colors=COLOR_NEG)
ax.set_ylabel('V / V vs $Li/Li^+$')
ax2.set_ylabel('V vs $Li/Li^+$')
ax.set_xlim(xlim)
ax2.set_ylim(neg_ylim)
ax.set_ylim(pos_ylim)
Comparison of metrics
df = pd.DataFrame([esoh_data_1, esoh_data_2], index=['1', '2'])
df = df.drop(columns=['ful', 'pos', 'neg', 'orig', 'Xt'])
# df = df.transpose()
df
| np_ratio | Cp | Cn | y100 | y0 | x0 | x100 | n_li | Cn_excess | Cp_excess | Cf | RMSE_mV | Cn_pf | x100_pf | cycle_index | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 0.998022 | 2.739773 | 2.734355 | 0.037883 | 0.950368 | 0.007241 | 0.921533 | 0.09789 | 0.32 | 0.14 | 2.50 | 21.222650 | 2.856612 | 0.889973 | 3.0 |
| 2 | 1.045994 | 2.619009 | 2.739467 | 0.031870 | 0.917702 | -0.000072 | 0.846808 | 0.08967 | 0.53 | 0.22 | 2.32 | 25.899812 | 2.797880 | 0.833661 | 159.0 |
Percentage changes:
df.pct_change(axis='rows')*100
| np_ratio | Cp | Cn | y100 | y0 | x0 | x100 | n_li | Cn_excess | Cp_excess | Cf | RMSE_mV | Cn_pf | x100_pf | cycle_index | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN | NaN |
| 2 | 4.806643 | -4.407812 | 0.186963 | -15.87224 | -3.437201 | -100.999138 | -8.108813 | -8.397637 | 65.625 | 57.142857 | -7.2 | 22.038538 | -2.056006 | -6.327446 | 5200.0 |
plt.figure(figsize=(13, 9))
ax1 = plt.subplot(121)
ax2 = plt.subplot(122)
for cellid in cellid_list:
print(f'Processing cell {cell.cellid}...')
cell = cell_dict[cellid]
if cell.cellid == 9:
continue
if cell.is_baseline_formation():
color = COLOR_BASE
else:
if cell.is_room_temp():
color = COLOR_FAST_RT
else:
color = COLOR_FAST_HT
if cell.is_room_temp():
ax = ax1
else:
ax = ax2
res = cell.calculate_var_q_c20_discharge()
ax.plot(res['var_q_c20_c56_c3_delta_q'], res['var_q_c20_c56_c3_voltage_v'], color=color)
ax1.set_ylabel('Voltage (V)')
ax1.set_xlabel('$\Delta Q$ (Ah)')
ax2.set_ylabel('Voltage (V)')
ax2.set_xlabel('$\Delta Q$ (Ah)')
ax1.set_title('Room Temperature')
ax2.set_title('45C')
Processing cell 40... Processing cell 1... Processing cell 2... Processing cell 3... Processing cell 4... Processing cell 5... Processing cell 6... Processing cell 7... Processing cell 8... Processing cell 9... Processing cell 10... Processing cell 11... Processing cell 12... Processing cell 13... Processing cell 14... Processing cell 15... Processing cell 16... Processing cell 17... Processing cell 18... Processing cell 19... Processing cell 20... Processing cell 21... Processing cell 22... Processing cell 23... Processing cell 24... Processing cell 25... Processing cell 26... Processing cell 27... Processing cell 28... Processing cell 29... Processing cell 30... Processing cell 31... Processing cell 32... Processing cell 33... Processing cell 34... Processing cell 35... Processing cell 36... Processing cell 37... Processing cell 38... Processing cell 39...
Text(0.5, 1.0, '45C')
